Songklanakarin J. Sci. Technol. 37 (6), 675-681, Nov. - Dec. 2015
http://www.sjst.psu.ac.th
Original Article
Application of 2D electrical resistivity tomography to engineering projects: Three case studies
Rungroj Arjwech1* and Mark E. Everett2
1 Department of Geotechnology, Faculty of Technology, Khon Kaen University, Mueang, Khon Kaen, 40002 Thailand.
2 Department of Geology and Geophysics, Texas A&M University, College Station, Texas.
Received: 2 December 2014; Accepted: 1 July 2015
Abstract
Electrical resistivity tomography (ERT) is a non-invasive geophysical method of primary interest for addressing subsur- face engineering problems. The method is based on the assumption that subsurface geological materials have significant resistivity contrasts that can be identified based on measurements on the surface. This paper presents three different case studies that have been carried out at different sites. The first case study visualizes the contrast between high resistivity zones of hard bedrocks and low resistivity zones of weathered rocks. Similar to the first case study, the second case study shows high resistivity contrasts that clearly distinguishes the shape of a footing located within the surrounding materials. The third case study shows no clear low resistivity zone that can be identified as a leaking zone. The 2D ERT survey method used in these three investigations has been shown to be useful as a cost-effective and rapid method to obtain wide area subsurface informa- tion that is relevant for subsurface engineering problems.
Keywords: non-invasive geophysical methods, electrical resistivity tomography (ERT), unknown bridge foundation determination, seepage
1. Introduction maintenance of highways (Dahlin, 2001; Wightman and Jalinoos, 2003). In many instances, geophysical NDT Geophysics involves the use of non-invasive tech- methods enhance the reliability and speed, and also reduce niques to determine subsurface anomalies without having to the cost of a geotechnical investigation (Anderson et al., 2008). engage in destructive excavation (Barker, 1993). Non-destruc- Assessing and characterizing geotechnical conditions can tive testing (NDT) is defined as the evaluation of the proper- become complex and costly in the presence of obstacles such ties of a material, component or system without causing as difficult access, irregular terrain and ground conditions, or damage (Louis, 1995). In the last few decades, geophysical regulatory constraints. Results based on traditional methods NDT methods have been developed and increasingly applied such as penetration testing or direct sampling may be of for addressing engineering problems. As one example, trans- limited utility. Surface geophysical techniques can provide portation personnel have used geophysical NDT methods alternate, wide-area methods for subsurface characterization in assisting geotechnical site investigation, construction, and and information regarding relevant material properties (Rucker, 2006). Though geophysics is not a substitute for geotechnical boring or testing, it is often a very cost-effective and efficient means of constructing contiguous 2D and 3D * Corresponding author. images of the subsurface and determining in-situ bulk Email address: [email protected] properties (Anderson et al., 2008). 676 R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015
The electrical resistivity tomography (ERT) method is one of the most widely used near-surface geophysical survey I 1 1 1 1 V , (1) methods for civil engineering applications (Castilho and Maia, 2 r1 r 2 r 3 r 4 2008). The method has been used for mapping electrical where r is resistivity, while r , r r and r are distances of resistivity in two and three dimensions (Dahlin, 2001). 1 2, 3, 4 the potential electrodes P and P from the current electrodes Previous studies have demonstrated the use of the ERT 1 2 C and C , respectively. Equation 1 is valid if the ground has method for identification of bedrock structures (Hsu et al., 1 2 homogeneous resistivity. 2010; Chambers et al., 2013), cavities or sinkholes (Kaufmann In case of inhomogeneous ground, an apparent resis- et al., 2012; Gómez-Ortiz and Martín-Crespo, 2012), geo- tivity r is calculated from the relationship between the technical site investigation (Al-Fares W., 2011; Haile and a applied current and the potential difference for a particular Atsbaha, 2014), slope stability investigation (Marescot et al., electrode arrangement and spacing. It is defined by, 2008; Perrone et al., 2014), and unknown bridge foundation determination (Arjwech et al., 2013; Tucker et al., 2014). k V , (2) The ERT method may be used for various other a I purposes in subsurface engineering investigations. This where k is a geometric factor dependent on the electrode paper aims to further demonstrate the application of the ERT spacing, technique on a number of engineering problems. More specifically, the paper presents the results of 2D ERT that 2 k . (3) have been carried out by the authors in various projects 1 1 1 1 including investigating the subsurface geology of a building r r r r construction site, determining the depth of an unknown 1 2 3 4 bridge foundation, and determining seepage from the earthen The apparent resistivity clearly depends on the geo- embankments of a wastewater treatment pond system. metry of the electrode configuration. The best electrode configuration for a field survey depends on the sensitivity 2. Electrical Resistivity Survey of the resistivity meter, the background noise level, and the relative importance assigned by the geophysicist to depth The electrical resistivity technique is based on the of penetration and lateral resolution. Standard electrode con- assumption that subsurface geological materials exhibit a figurations used for 2D ERT surveys are Wenner, dipole- wide variability of resistivity values and that geological dipole, Wenner-Schlumberger, pole-pole, pole-dipole, and boundaries can be identified based on measurements of resis- equatorial dipole-dipole (Telford et al., 1990; Loke, 2000; tivity. If a target of interest has a sufficiently large electrical Kearey and Brooks, 2002; Dahlin and Zhou, 2004; Loke and resistivity contrast with respect to that of the surrounding Lane, 2004; Loke, 2010). material, it can be detected by surface measurements of voltage following the injection of current through pairs of 3. 2D Electrical Resistivity Tomography electrodes (Barker, 1993). The purpose of a resistivity survey is thus to deter- A 2D multi-electrode ERT survey may be carried out mine the distribution of underground resistivity from using a large number of electrodes connected to a multi-core measurements of potential difference, or voltage, made on the cable. The electrode cable is typically divided into sections of ground surface. An electric current I (amperes, A) is injected manageable length, which are then connected end-to-end. at electrode C1 and withdrawn at electrode C2 as shown in Electrodes connected to the cable take-outs are inserted into
Figure 1, while two other electrodes P1 and P2 are used to the ground at a specified regular interval along a survey line. record the resulting potential difference ÄV (volt, V), A resistivity meter and electronic switching unit are used in conjunction with a user-programmed protocol to automati-
cally measure ra in a pre-defined sequence of combinations of four electrodes. Efficient data acquisition is achieved by measuring several voltages simultaneously across multiple pairs of electrodes following a single injection of electric current (Loke, 2000; Bernard, 2003; Hiltunen and Roth, 2003; Loke, 2010). Figure 2 shows an example of electrode arrangement and measurement sequence for a 2D ERT survey. When the data acquisition is completed, data analysis is performed Figure 1. Two current (C1 and C2) and two potential (P1 and P2) electrodes in the standard configuration (Telford et al., using the RES2DINV (Loke, 2004) software, including 2D 1990). pseudo-section plotting, and inversion. The RES2DINV R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015 677
Figure 2. Arrangement of electrodes for a 2D survey and the sequence of measurements used to build up an apparent resistivity pseudo-section (Loke, 2000). inversion algorithm is described by Loke and Barker (1995; 2 show similar resistive zones and thicknesses (Figure 4). 1996) and Yang (1999) and is based on a smoothness- The prominent low-resistivity zone of <100 m in the near- constrained least squares approach (DeGroot-Hedlin and surface towards the west corresponds to location without Constable, 1990). exposed bedrock and hence is interpreted as a zone of weathered rock and top-soil. The high-resistivity zone >200 4. 2D ERT to Investigate Subsurface Geology at a Building m that is dominant at the east end of the profiles is inter- Construction Site preted as sandstone bedrock. The inversion results from the profiles ChRU3 and 4 4.1 Site description show near-surface high resistivity >200 m which is consis- tent with resistant sandstone bedrock exposed on the A nine-story building is planned to be constructed surface. The high resistivity zone thins toward the north and on a site that is situated on a ridge characterized by cuesta extends to a maximum of ~15 m in depth at the southern ends topography aligned East-West (Figure 3A). The terrain is of the profiles. A low resistivity zone close to the northern covered with trees and bushes, with numerous boulders and ends can be seen at lower depths in both sections. These bedrock also contributing to make access difficult for a resis- lower zones of resistivity <100 m are the ones imaged on the tivity survey. The bedrock is well exposed on the surface and orthogonal profiles ChRU1 and 2. The inversion results of delineates the rim of a sedimentary basin structure. the profiles ChRU5 and 6 show similar structures as found in profiles ChRU3 and 4 but thicker high-resistivity zones are 4.2 Methods evident.
The objective of this study is to identify bedrock. ERT 4.4 Data verification data acquisition comprises six profiles (ChRU 1-6) using SYSCAL R1 Plus by IRIS Instrument. Hybrid Wenner- Due to a lack of well log and other subsurface data, Schlumberger electrode array configurations were selected the resistivity images were verified by comparing against with 5 m electrode spacing, yielding a total length of each available geological information nearby the site. A sandstone profile of 235 m. In order to cover the entire proposed site outcrop is well exposed on the front slope in Figure 5 and using only a few ERT profiles, the first two profiles were reveals clear stratification. The lithology consists of alternat- separated by 17.5 m and oriented East-West, whereas the ing beds of hard well-lithified sandstone overlying highly other four profiles were separated by 20 m intervals and weathered sandstone and siltstone with a distinctive sharp, oriented North-South (Figure 3D). erosive, and conformable contact. The zone of high resistivity >200 m in Figure 4 is interpreted to be caused by the resis- 4.3 Interpretation tant bed of sandstone, whereas the zone of lower resistivity <200 m is interpreted to be caused by the weathered The inversion results indicate that good resistivity sandstone and siltstone. The construction activities planned data were acquired, converged with a RMS misfit of lower for the site can take advantage of the subsurface information than 7 at the maximum fifth iteration. The profiles ChRU1 and that is provided by these ERT results. 678 R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015
Figure 3. View of (A) the cuesta topography of the study area; views (B and C) along survey profiles illustrating the complex terrain consisting of large boulders and bedrocks on the surface; schematic plan view of location of the resistivity profiles ChRU1-6 (D).
Figure 4. Resistivity fence diagram from all inversion images shows subsurface features of the construction site study area. Red and green colors correspond to high resistivity, whereas blue is low resistivity.
5. 2D ERT for Unknown Bridge Foundation Depth Determi- nation
5.1 Site description
A railway bridge built across a river has been identified as containing unknown foundations due to non-existent information about their design and construction. A represen- tative foundation located on the steep slope of the river bank is difficult to characterize using traditional exploratory Figure 5. Mapping of the geology from an outcrop study near the methods. For example, the water level rises and floods part of construction site shows a resistant sandstone bed on the the foundation during the rainy season. The foundation has top, overlying weathered sandstone and siltstone, with a hexagonal cross-sectional shape with 8, 3, and 3 m side sharp contact between the two units. lengths (Figure 6). R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015 679
5.3 Interpretation
Both (RWB1-2) inversions converged with a RMS misfit of lower than 4 after five iterations. The results indicate a strong contrast between the resistivity of the foundation and that of the surrounding geological materials, as shown in Figure 7. The ERT images generally show lower resistivity zones corresponding to geological materials and a central higher resistivity zone corresponding to the concrete founda- tion itself. The zones of low resistivity extend to both end Figure 6. ERT profile RWB1 was aligned parallel to the river (A); sections of the profile where clay particles and elevated a 3D schematic view of the railway bridge site showing moisture act to increase electrical conductivity. The shallow- resistivity survey profile and location of the hexagonal est exposed bedrock is found close to the foundation, with foundation (B). resistivity values ranging between 10-40 m. This layer represents weathered to moderately weathered shale, as observed on the surface. A high resistivity anomaly >80 m 5.2 Methods coincides with the concrete foundation. Its shape is some- what rectangular, being 11x5 m for RWB1 and 3x5 m for Unknown bridge foundations pose a significant safety RWB2, respectively and a depth of 5 m. This zone is inter- risk due to the possibility of their undermining by stream preted to be the resistivity signature of a large spread footing. scour and erosion. So the objective of this study is to visual- ize unknown foundation shape and depth. A 2D ERT data set 5.4 Data verification was collected for this study using the SuperStingTM R8/IP system by Advanced Geosciences Inc., (AGI). The survey Both resistivity images at the railway bridge site show consisted of two orthogonal profiles (RWB1-2) conducted that the horizontal size of the anomaly associated with the with dipole-dipole electrode configurations of 28-electrodes spread footing is consistent with the actual size of the at 2 m spacing. The total length of each survey profile is thus foundation (11x3 m wide). A bridge layout plan showing the 54 m. This layout provides the capability to map the narrow designed depth of the footing is not available so the ERT- vertical subsurface foundation. The profiles passed within interpreted foundation depth could not be verified. Without 0.5 m alongside the foundation and were laid out parallel confirmation documentation provided by the original layout and perpendicular to the river, respectively. The profiles plans, acquiring two surveys conducted on perpendicular intersected near one corner of the foundation, as shown in profiles increases the reliability relative to that provided by the Figure 6. a single profile in any direction.
Figure 7. Inversion results at the railway bridge clearly show a high resistivity zone interpreted as a shallow footing at about the midpoint of the survey profiles. 680 R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015
Generally, the 2D ERT method for foundation determi- The two survey profiles intersected at the southwest corner nation is somewhat straightforward to interpret due to the of the pond system. known location of the foundation along the profile. A detailed interpretation is usually done by qualitative comparison of 6.3 Interpretation the observed surface location with that inferred on the specific inversion image. The depth of the foundation can The inversion results were converged to a maximum of then be directly visualized on the inversion result. In other 5% of RMS error in five iterations, as shown in Figure 9. A low types of engineering problems, the horizontal location of the resistivity zone of < 20 Wm on profile WTPS1 is evident at target may not be known in advance. three distinct locations between distances 280-320 m near the surface on the downstream side. These zones are attributed to 6. 2D ERT for Determining Seepage of Earthen Embank- the effect of water running through known drainage systems. ments of a Wastewater Treatment Pond System The resistivity anomalies are consistent with their locations on the surface where the treated water is drained. However 6.1 Site description the low resistivity zones are distorted compared to the actual shape of the drains as expected for geophysical imaging. The A wastewater treatment pond system is constructed two inversion images show no other signs of low resistivity on Quaternary loess deposits. The pond system consists of zone identified as seepage anomaly. A low resistivity level multiple ponds in series (Figure 8). The concern of seepage located below 15 m is clearly seen. This zone is interpreted through the earthen embankments is rising because loess is as the underlying groundwater level. easily erodible and collapsible. Dry loess usually has high shearing resistance; however when wet it loses considerable 7. Conclusions shear strength (Phien-wej et al., 1991). The concern of geotechnical engineers about un- 6.2 Methods certainties in ground conditions suggests the utilization of 2D
The objective of this study is to image the possibility of seepage zone. Two 2D ERT profiles (WTPS1-2) were acquired along the west and south sides of the pond system using SYSCAL R1 Plus by IRIS Instruments. Possible seep- age is located on these sides because groundwater flow along the hydraulic gradient intersects them. The ERT survey profiles were deployed atop the earthen embankments using hybrid Wenner-Schlumberger electrode configurations with 48 electrodes and 5 m spacing. An additional “roll along” Figure 8. ERT profile WPTS2 was aligned North-South (A); sche- technique was adopted to extend of the length of each survey matic plan view of wastewater treatment pond systems profile. The length of WTPS1 is 445 m and WTPS2 is 390 m. showing orthogonal resistivity survey profiles (B).
Figure 9. Inversion images show no indication seepage. The low resistivity zones are interpreted as drainage systems that are consistent with their known locations on surface. R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015 681
ERT for characterizing the subsurface. The three case studies Chambers, J.E., Wilkinson, P.B., Wardrop, D., Hameed, A., presented herein demonstrate the successful use of 2D ERT. Hill, I., Jeffrey, C., Loke, M.H., Meldrum, P.I., Kuras, The first study showed that ERT can be used for planning O., Cave, M. and Gunn, D.A. 2012. Bedrock detection building construction, as it is a technique that can discrimi- beneath river terrace deposits using 3D electrical nate between resistant and weathered bedrocks. ERT is also resistivity tomography. Geomorphology. 177-178, 17- seen to be an effective tool for imaging the depth of large- 25. shallow bridge foundations. Finally, ERT can be used effec- Dahlin, T. 2001. The development of dc resistivity imaging tively to determine whether there is seepage through earthen techniques. Computers and Geosciences. 27, 1019- embankments of wastewater treatment pond systems. 1029. The 2D ERT method used in these three investiga- Dahlin, T. and Zhou, B. 2004. A numerical comparison of 2D tions is as a cost-effective and rapid means to obtain wide resistivity imaging with 10 electrode arrays. Geophysi- area subsurface information. Though 2D ERT is not a substi- cal Prospecting. 52, 379-398. tute for geotechnical boring or testing, it is a non-invasive DeGroot-Hedlin, C. and Constable, S. 1990. Occam’s inversion technique, combining rapid acquisition, and safe operation. to generate smooth, Two-dimensional models form The equipment is portable and setup can often be effectively magnetotelluric data. Geophysics. 55, 1613-1624. deployed over densely vegetated or steep slope areas that Gómez-Ortiz, D. and Martín-Crespo, T. 2012. Assessing the might not be easily accessible to traditional invasive risk of subsidence of a sinkhole collapse using ground methods. The ERT results provide 2D subsurface images with penetrating radar and electrical resistivity tomography. good spatial resolution along the survey profile. Based on the Engineering Geology. 149–150, 1–12. experience gained by applications such as three case studies Haile, T. and Atsbaha, S. 2014. Electrical resistivity tomo- presented herein, it is recommended that the 2D ERT method graphy, VES and magnetic surveys for dam site should be increasingly used by geotechnical engineers. characterization, Wukro, Northern Ethiopia. Journal Moreover, other complementary geophysical methods can be of African Earth Sciences. 97, 67–77. used in an integrated exploration that may be able to enhance Hiltunen, D. and Roth, M. 2003. Investigation of bridge images of subsurface geological materials over images foundation sites in karst terrain via multi-electrode obtained using the ERT method alone. electrical resistivity, Proceeding of the 3rd International Conference on Applied Geophysics – Geophysics, Acknowledgements 2003. Hsu, H., Yanites, B., Chen, C. and Chen, Y. 2010. Bedrock We gratefully acknowledge the Texas Department of detection using 2D electrical resistivity imaging along Transportation, Texas, U.S.A. and the Department of Geo- the Peikang River, central Taiwan. Geomorphology. technology, Khon Kaen University, Thailand, for support 114, 406-414. and encouragement. Kaufmann, O., Deceuster, J. and Quinif, Y. 2012. An electrical resistivity imaging-based strategy to enable site-scale References planning over covered palaeokarst features in the Tournaisis area (Belgium). Engineering Geology. 133- Al-Fares, W. 2011. Contribution of the geophysical methods 134, 49-65. in characterizing the water leakage in Afamia B dam, Kearey, P. and Brooks, M. 2002. An Introduction to Geo- Syria. Journal of Applied Geophysics. 75, 464-471. physical Exploration, Blackwell Science Publications, Anderson, N., Hoover, R. and Sirles, P. 2008. Geophysical U.S.A. methods commonly employed for geotechnical site Loke, M.H. and Baker, R.D. 1995. Least-square deconvolu- characterization. Transportation Research Broad of tion of apparent resistivity pseudosections. Geo- the National Academies. physics. 60, 1682-1690. Arjwech, R., Everett, M., Briaud, J.L., Hurlebaus, S., Medina- Loke, M.H. and Barker, R.D. 1996. Practical techniques for 3D Cetina, Z., Tucker, S. and Yosefpour, N. 2013. Electrical resistivity surveys and data inversion. Geophysical resistivity imaging of unknown bridge foundations. Prospecting. 44, 499-523. Near Surface Geophysics. 11, 591-598. Loke, M.H. 2000. Electrical Imaging s Surveys for Environ- Barker, P. 1993. Techniques of archaeological excavation. mental and Engineering Studies, a practical guide to Taylor and Francis e-Library. 2-d and 3-d surveys. Penang, Malaysia. Bernard, J. 2003. The depth of investigation of electrical Loke, M.H. 2004. RES2DINV ver. 2.14, Geotomo Software Sdn methods. IRIS-instruments, France. Bsd, Malaysia. Castilho, G. and Maia, D. 2008. A successful mixed land Loke, M.H. and Lane, J.W. 2004. Inversion of data from underwater 3D resistivity survey in an extremely electrical resistivity imaging surveys in water covered challenging environment in Amazônia. 21th EEGS areas. Exploration Geophysics. 35, 266-271. Symposium on the Application of Geophysics to Engi- Loke, M.H. 2010. 2D and 3D Electric Imaging Surveys, Geo- neering and Environmental Problems. 1150-1158. tomo Software Sdn Bsd, Malaysia. 682 R. Arjwech & M. E. Everett / Songklanakarin J. Sci. Technol. 37 (6), 675-681, 2015
Louis, C. 1995. Nondestructive Testing; radiography, ultra- Telford, W.M., Geldart, L.P. and Sheriff, R.E. 1990. Applied sonics, liquid penetrant, magnetic particle, eddy Geophysics, Cambridge University Press, U.K. current, ASM International. Materials Park, Ohio, Tucker, S., Hurlebaus, S., Everett, M., Briaud, J.L., Medina- U.S.A. Cetina, Z., Yousefpour, N. and Arjwech, R. 2014. Marescot, L., Régis, M. and Chapellier, D. 2008. Resistivity Electrical resistivity and induced polarization imaging and induced polarization surveys for slope instability for unknown bridge foundations. Journal of Geo- studies in the Swiss Alps. Engineering Geology. 98, technical and Geoenvironmental Engineering. 141(5), 18–28. 040150081-0401500811. Perrone, A., Lapenna, V. and Piscitelli, S. 2014. Electrical resis- Wightman, W. and Jalinoos, F. 2003. Application Geophysical tivity tomography technique for landslide investiga- Methods to Highway Related Problem, Federal High- tion. Earth-Science Reviews. 135, 65-82. way Administration, U.S. Department of Transporta- Phien-wej, N., Pientong, T. and Balasubramaniam, A.S. 1991. tion, U.S.A. Collapse and strength characteristics of loess in Thai- Yang, X. 1999. Stochastic Inversion of 3D ERT Data, Ph.D. land. Engineering Geology. 32, 59-72. Thesis, University of Arizona, U.S.A. Rucker, M. 2006. Surface geophysics as a tool for charac- terization existing bridge foundation and scour condi- tions. Proceedings of the Conference on Applied Geophysics, Missouri, U.S.A., December 4-7, 2006.